Survival to recovery in children with severe acute malnutrition treated at outpatient therapeutic care program in Southern Ethiopia

Abstract

Background: Community-based management of severe acute malnutrition has been widely rolled out and integrated to the existing health care system in Ethiopia. But there is a dearth of information concerning the survival to recovery of severely acutely malnourished children after rolling out of the program to a health post level in the country. Objectives: To examine survival to recovery in children with severe acute malnutrition treated at outpatient therapeutic care program in Southern Ethiopia. Method and materials: Institution based retrospective cohort study was conducted using a data available from Jan. 2011 to Jan. 2013 in twelve randomly selected health posts. A two population proportion formula was used to calculate a sample size of 374. Descriptive analysis was done using percentages for categorical data and mean/median for continuous variables. Chisquare was used to measure associations between categorical variables. The Mann-Whitney U test for the independent two samples test was used to compare the medians for non-normally distributed continuous variables. The outcome variable was time to recovery. The life table analysis and the Kaplan-Meier product limit were used to estimate the survival characteristics of the study subjects. And the log rank test was used to compare the survival curves. The Cox proportional-hazard regression model was used determine predictors of time to recovery. Result: A total of 348 patient cards were reviewed. The median time to recovery was 35 days and 49 days for children with kwashiorkor and marasmus respectively. It was significantly different (Log Rank=46.93, df=1 P<0.001). There was a 1.2% decrease in the likelihood of recovery for one month increase in age (AHR=0.988, (0.977, 0.999)). Children with marasmus were also 48% less likely to recover (AHR=0.517, 95% CI (0.386, 0.691)). Children who had a weight gain of >=3.21gm/Kg/d were 2.43 times (AHR=2.434, (1.828, 3.241)) more likely to recover. For a MUAC gain (mm/day) of one unit increase there was a 2.33 times (AHR=2.326, (1.373, 3.942)) increase in the likelihood of recovery. Conclusion: Being marasmic, age, weight and MUAC gain were identified as a predictor of recovery. Hence it needs special attention on counseling the care taker about the appropriate way of feeding the therapeutic diet to the sick child to improve weight and MUAC gain with due emphasis given to marasmic children.